The central goal of this proposal is to define, at the level of nucleotide sequences of the genes, the mutations that cause various forms of epidermolysis bullosa (EB). Several different strategies will be used to focus on the elucidation of the type of mutations that are most accessible for evaluation by current recombinant DNA technologies. We will develop these strategies in such a way that they will generate as much new information on EB as possible. These technologies will also be of use in studying other genetic diseases affecting the skin. The principal strategy is to select candidate genes in different forms of EB, on the basis of the following criteria: (a) evidence from genetic linkage studies in Project 1, indicating that a specific allele for a gene is co-inherited with the disease; (b) evidence from morphological or immunofluorescence studies, performed and part of the diagnostic evaluation of the patients under Core C, suggesting that the gene for a specific protein is aberrantly expressed, or that the protein synthesized from the gene is structurally or metabolically abnormal. Based on these criteria, we have definitely identified type VII collagen as a candidate gene in four families with the dominant dystrophic form of EB. Furthermore, genetic linkage of an EB simplex mutation in one family to a locus adjacent to the keratin type II gene cluster on chromosome 12q suggests that some of the keratins (such as keratin 5) expressed in the basal layer of the epidermis, are candidate genes in this family. In these cases, we will focus our efforts on the candidate genes using gene probes that have been and will be developed or are available from other investigators. In each case, the specific aim will be to determine the nucleotide sequence in and around the site of the mutation.